EXTREMELY VIOLENT NEWS IMAGES

EXTREMELY VIOLENT NEWS IMAGES

Sometimes television networks or stations or print media are faced with a difficult decision about whether to air a graphically violent news photograph. One of the most agonizing decisions came in July 2003 after the deaths of deposed Iraqi dictator Saddam Hussein’s sons Oday and Qusay in an assault by the U.S military on a fortified house in Mosul, Iraq. At this point, Iraq was occupied by U.S. and U.K. forces but Saddam was apparently still alive and in hiding. The two sons were notoriously hated and feared for perpetrating crimes of incredible cruelty. Although the United States generally had a policy of not releasing photographs of people killed in military action, it decided that at least the Iraqi public needed concrete evidence that the hated brothers were really dead. Thus the gruesome photos of the dead brothers were published.

Sometimes cameras happen to capture an unintended gruesome image. Local media photographers covering a routine 1987 news conference by Pennsylvania Treasurer C. Budd Dwyer captured the unexpected image of Dwyer putting a pistol in his mouth and pulling the trigger, killing himself instantly. Network news, TV stations, and newspapers then faced the decision of whether to run the grisly sequence. Most TV stations chose not to air the tape or did so only to the point of Dwyer placing the gun in his mouth A few TV stations and newspapers carried pictures of

Violence: Watching All That Mayhem Really Matters 270

what happened after that, saying it was “an important historical event” and should be covered. In a 1993 case, a Florida woman being interviewed by a television crew in a park in broad daylight was suddenly accosted and shot several times by her estranged husband. Although the unexpected murder photos later provided important legal evidence, networks were faced with the decision of whether or not to air the footage on the news. Many (including CNN) did so.

The public has a right to know, but how much does the public have a need or right to know and see explicitly? The answer is obviously an ethical and policy question, although one study suggested a positive effect of learning at least some gory details. College students who read a newspaper report of a wife-battering incident rated it more seriously and rated the batterer more negatively if the victim’s injuries were described in detail, compared to a situation in which they were not explicitly described (Pierce & Harris, 1993). These descriptions, however, were much less graphic than the pictorial images described above.

BOX 9.4

HOLLYWOOD GOES TO VIETNAM The 1966 pro-military film The Green Berets starred John Wayne in a

stylized epic with good guys and bad guys. By 2 or 3 years later, such a simplistic approach to Vietnam rang very shallow and false. A very few years after The Green Berets, sentiment in the United States had turned completely around, and whatever glory there had ever been to the Vietnam War had rapidly disappeared. The 1974 Oscar-winning documentary Hearts and Minds, by Peter Davis, was very graphic and carried such a strong antiwar message that some newspapers refused to publish reviews of the film. Although the last U.S. troops left Vietnam in 1973 and Saigon fell in 1975, there were practically no commercially successful film images of Vietnam before 1978, when The Deer Hunter and Coming Home picked up seven Oscars between them. Both of these (along with Apocalypse Now in 1979) carried a strong antiwar message and reflected the country’s disgust with the war in Vietnam, a wound that had just started to heal.

That healing was not complete in 1986, when Oliver Stone’s graphic and realistic Platoon and Sylvester Stallone’s live-action comic book Rambo: First Blood Part II appeared. Both were huge successes, but Platoon was the greater commercial gamble, having taken years to acquire the necessary funding The Pentagon refused to help the

271 A Cognitive Psychology of Mass Communication

producers, although it enthusiastically aided the makers of the heroic Top Gun, on the grounds that it wanted to ensure a realistic portrayal of the military! Widely hailed as a critical success (including winning the Best Picture Oscar), Platoon astounded everyone with its huge commercial success as well. It was then followed by a spate of realistic Vietnam films like Gardens of Stone, Casualties of War, The Hanoi Hilton, and Full Metal Jacket, none of which received close to Platoon’s critical or commercial success.

By 1988, the wounds of Vietnam had healed enough to allow the production and modest commercial success of the first Vietnam War TV series (Tour of Duty) and big-screen comedy Good Morning, Vietnam. A comedy about Vietnam would have been unthinkable much before that time. Not too long later followed another Oliver Stone antiwar film, Born on the Fourth of July, the biographical story of gung-ho soldier turned antiwar protester Ron Kovic.

Although sensitization effects are hard to study scientifically for ethical reasons, general sensitization effects are probably not too widespread and do not occur nearly as often as their opposite, desensitization (see below). In general, situations for which one can posit sensitization effects can often be interpreted equally plausibly in terms of desensitization. For example, people have argued that daily news broadcasts of the Vietnam War sensitized us to the horrors of war and eroded public support for that conflict, in contrast to previous wars. On the other hand, others have argued that the same news broadcasts desensitized us to war, and thus we now are not as bothered by seeing images of other conflicts (see chapter 7). See Box 9.4 for a history of the treatment of the Vietnam War in films.

Desensitization

Although public debate about the effects of TV violence typically worries primarily about increases in violent behavior, there may be far more pervasive attitudinal effects, especially in the area of desensitization. The basic principle here is that viewing a steady diet of violence in the media makes us less sensitive to it, more jaded, and less aroused and bothered by it. We become so used to seeing people wasted, blown apart, or impaled that it no longer particularly troubles us. For example, after seeing a violent TV show, sixth graders were less sensitive to violent images in a subsequent film than were children who had seen a nonviolent film first (Rabinovitch, McLean, Markham, & Talbott, 1972), Desensitization is typically measured experimentally through physiological and/or attitudinal measures.

Violence: Watching All That Mayhem Really Matters 272

How Desensitization Works. Desensitization may be seen as a straightforward example of classical conditioning (see Fig. 9.1). The normal, unlearned responses to being physically hurt include pain, fear, and disgust. The first time one sees media violence, it probably evokes such negative emotional responses, due to its similarity to real violence (Fig. 9.1 a). Such a single occurrence may actually produce sensitization as discussed above.

FIG. 9.1. Desensitization as classical conditioning. What happens with repeated viewing of violence in comfortable

surroundings is quite different, however (Fig. 9.1b). Suppose, for example, that the normal, unlearned response to sitting at home in one’s easy chair is feeling relaxed and happy. When this is repeatedly paired with violence on TV and movies, vicarious violence in a pleasant home context gradually becomes associated with that situation and itself comes to be seen as entertaining, pleasant, and even relaxing. The natural association of filmed violence and real-life violence has been weakened as the new association of video violence with recreation is strengthened. We repeatedly see violence without experiencing pain or hurt ourselves and thus the normal negative responses to it weaken. Given what we know about classical conditioning in psychology, it is unlikely that such frequent and repeated exposure to stimuli could fail to have a substantial effect. In the adolescent subculture, part of the male gender-role socialization has been for a boy to desensitize himself

273 A Cognitive Psychology of Mass Communication

so that he can watch graphic violence and not appear to be bothered by it (Harris, et al., 2000; Mundorf, Weaver, & Zillmann, 1989; Tamborini, 1991; Zillmann & Weaver, 1996; Zillmann, et al., 1986). Demonstrating desensitization to violence thus becomes a way to impress a date.

Consequences of Desensitization. What are the implications of people becoming desensitized to violence from the media? Becoming jaded to news of war and violence will cause such stories not to bother us so much anymore. Even if we never come to actually like violence or behave violently ourselves, we may not dislike it nearly so much; it does not seem all that serious. This has important implications for behavior. For example, Drabman and Thomas (1974, 1976) had 8- to 10-year-old children watch a violent or nonviolent film and later watch younger children at play. When the younger children started to get rough, the older children who had watched the nonviolent film called an adult sooner than did the older children who had watched the violent film, thus showing some generalization of the desensitization effect.

One of the major areas of concern with desensitization has to do with tolerance of violence toward women. Male college students who viewed a series of slasher horror movies later showed less empathy and concern for victims of rape (Linz, Donnerstein, & Penrod, 1984). Sexual violence is one of the major current concerns among media researchers studying violence; we examine this aspect of violence in some detail in chapter 10.

Cultivation

Another type of attitudinal effect in regard to violence is cultivation. As discussed in chapter 2, George Gerbner and his colleagues argue that the more exposure a person has to television, the more that person’s perception of social realities will match what is presented on TV (Gerbner, Gross, Morgan, Signorielli, & Shanahan, 2002; Signorielli & Morgan, 1990; Weaver & Wakshlag, 1986). In fact, cultivation theory was first developed in regard to studying media violence. In contrast to modeling, cultivation attributes a more active role to the viewer, who is interacting with the medium, not being passively manipulated by it. Nevertheless, there is a coming together of the viewer’s outlook and that of the medium, whereby the person’s perceived reality gradually approaches that of the TV world.

Cultivation theory is best known for its research on the cultivation of attitudes related to violence (Gerbner, Gross, Morgan, & Signorielli, 1980; Gerbner, Gross, Signorielli, & Morgan, 1986). Such studies show that frequent viewers believe the world to be a more dangerous and crime-ridden place than infrequent viewers believe it to be. The world of TV shows 50% of characters involved in violence each week, compared to less than 1% of

Violence: Watching All That Mayhem Really Matters 274

the population per year in real life. This cultivation effect could be due either to TV teaching that this is what the world is like or to the fact that more fearful people are drawn to watching more TV If it is the former, and cultivation theorists believe it is, TV can induce a general mindset about the position of violence in the world, completely aside from any effects it might have in teaching violent behavior. Finally, cultivation theory speaks of television teaching the role of the victim. From watching a heavy diet of crime and action-adventure shows, viewers learn what it is like to be a victim of violence, and this role becomes very real to them, even if it is completely outside of their own experience.

The effects of media violence just discussed are not presented as an exhaustive list but rather as general classes into which most proposed effects fall. Occasionally an effect falls outside of those classes, however; see Box

9.5 for some interesting evidence of violent media causing amnesia in viewers.

Important Interactive Factors

Now that we have looked at the major effects of watching media violence, it is time to examine the major moderating variables that affect how much the violent behavior will be modeled or the attitudinal effects of desensitization, cultivation, or fear will be induced. Consistent with the conditional effects model based on individual differences discussed in chapter 2, a violent model does not affect everyone the same way under all circumstances. Numerous important variables heighten or attenuate effects like modeling or desensitization. Violent media do not affect everyone in all circumstances in the same way.

BOX 9.5

DO VIOLENT IMAGES CAUSE AMNESIA? Amnesia is not on the typical lists of effects of media violence, but just

such an effect has been proposed (Christianson & Loftus, 1987; Loftus & Burns, 1982; Newhagen & Reeves, 1992). It has been known for some time that a physical injury to the brain can result in a loss of memory for events immediately preceding the impact. For example, the shock of one’s head flying against the headrest in an auto collision may lead to amnesia for events immediately preceding the impact. It is as if the brain had not yet had time to transfer the event from short-term to long-term memory. Loftus and Burns (1982) demonstrated that such an effect may also occur solely from the mental shock of seeing graphic violence on the screen Research participants saw a 2 minute film of a bank robbery in

275 A Cognitive Psychology of Mass Communication

either a violent or nonviolent version. In the violent version the fleeing robbers shot their pursuers and hit a young boy in the street in the face, after which he fell, clutching his bloody face. The nonviolent version was identical up to the point of the shooting, at which point the camera cut to the interior of the bank. Measured using both recall and recognition measures, people seeing the nonviolent version of the film remembered the number on the boy’s t-shirt better than those seeing the violent film, although the shirt was shown the same amount of time in both. A second study ruled out the possibility that the effect could have been due to the unexpectedness or surprisingness of the shooting.

Model Attributes. First, several characteristics of the violent model are important. People are more likely to imitate or be disinhibited by viewing the violent behavior of an attractive, respected, prestigious model than by one who does not have such qualities. Also, the more deeply we identify and empathize with a model, the more likely we are to imitate that person (Huesmann, Lagerspetz, & Eron, 1984; Huesmann, Moise-Titus, Podolski, & Eron, 2003). These points suggest that violence by the characters we admire and identify with is a stronger influence than violence by the bad guys. This has important ramifications for assessing effects of action-adventure and police shows.

Reinforcement of Violence. Whether or not the violence is reinforced in the plot is also a very important moderating variable. One of the central principles of operant conditioning, and indeed of all psychology, reinforcement refers to any event that follows a response and increases the probability of that response occurring again. The connection (contingency) between the response and the reinforcement is learned; thus the response is made in anticipation of receiving the reinforcement. A dog learns to fetch a stick because he is reinforced with a dog biscuit when he performs the act. A girl does her homework each night because she is reinforced by being allowed to watch TV after she is finished. After learning has occurred, responses continue to be made for some period of time without the reinforcement, until they gradually diminish and are finally extinguished altogether.

If acting violently appears to pay off for the violent character (in money, power, relationship, etc.), it is thus reinforced in the context of the story. Research suggests that reinforced violence is more likely to desensitize or be modeled than nonreinforced or punished violence (e.g., Bandura, 1965). Krcmar and Cooke (2001) found that 4- to 7-year-olds thought unpunished violence in a video clip was more right than punished violence. In a typical TV story line, violence by the hero is more likely to be rewarded (reinforced)

Violence: Watching All That Mayhem Really Matters 276

than is the violence of the villain, although the latter may have been reinforced for much of the show. Model characteristics and reinforcement suggest that a particularly troubling type of violent perpetrator is the child or teen. Content analyses show that, compared to adult models, child violent models in media are more attractive and less likely to be punished or experience other negative consequences of the violence (Wilson, Colvin, & Smith, 2002); both of these characteristics make modeling by the viewer and other effects more likely.

Finally, media, rather than reinforcing behavior or tendencies to behave in certain ways, may reinforce certain values about the use of violence. For example, characters on action-adventure TV shows and movies frequently use violence to settle interpersonal disputes. As such, they are subtly reinforcing the value that violent behavior is a realistic and morally acceptable manner of dealing with conflict, a value that may become part of the viewer’s perceived reality. Children seeing a fantasy movie clip ending in violence later rated a different story with violence as more morally correct than children seeing the same clip but without the violence (Krcmar & Curtis, 2003). When sportscasters legitimatize sports violence as necessary or regrettably acceptable in the context in which it occurred, or when an athlete receives only a slap on the wrist for assaulting another player, such treatment reinforces violence as a way of dealing with the stresses of the game.

Perceived Reality. Another important moderating factor is whether the violence is seen as real or make-believe, that is, the degree of perceived reality (van der Voort, 1986). There is some evidence of stronger effects of violence that is perceived as real than of violence that is perceived as unreal. Although children’s cartoons are by far the most violent genre of TV show, the violence is also the most stylized and unrealistic. Some studies (e.g., Feshbach, 1976) show cartoon violence to have less negative effect than more realistic violence.

In understanding the perceived reality of violent television, it is important to consider the child’s cognitive understanding of television at any given time (e.g., Cantor, 1998b, 2002; Cantor & Sparks, 1984; Gunter, 1985; Sparks, 1986). A very young child might think that a violent death on a police drama actually shows someone dying, rather than merely an actor pretending to die. Children who believe such staged violence to be real are often more disturbed by it than those who understand the convention of acting. The greater the perceived realism of media violence, the more likely that aggressive behavior will increase (Huesmann, et al., 2003). Continuing this line of reasoning, the most difficult forms of TV violence for children to deal with are probably news and documentaries, because violence on these programs is real and not staged. Beyond the issue of perceived realism, the

277 A Cognitive Psychology of Mass Communication

whole area of how the viewer interprets the violence is very important; such variables often account for more of the variance than stimulus factors manipulated by the experimenter (Potter & Tomasello, 2003).

Personality Characteristics of the Viewer. Research has also generally found larger modeling and other effects in people more naturally inclined toward violence in terms of personality (Heller & Polsky, 1975; Parke, et al., 1977), although this result has not been found consistently (e.g., see Huesmann, Eron, Lefkowitz, & Walder, 1984), Violence in the media may reinforce dispositional violent tendencies already present in the viewer, although it is not the cause of those tendencies. The more that such tendencies are reinforced, the more likely they are to manifest themselves in behavior. Regrettably, the lack of uniform effect on viewers has often been used to argue that there is no substantial impact of media violence. As suggested previously, however, a modeling effect on even a tiny percentage of the population may be cause for serious concern. The fact that people less naturally inclined toward violence do not respond as strongly as those more inclined toward violence is not an argument that media violence has no effect!

Another important factor in moderating the effects of media violence perceived by the viewer may be the thoughts that one has in response to viewing media violence (Berkowitz, 1984). Those thoughts may focus on the suffering of the victim, the triumph of the violent person, the relation of the violence to one’s own experience, and so on. Depending on the nature of these thoughts, their mediating role in facilitating violent behavior may vary substantially. For example, someone who is overwhelmed by the suffering of the victim is probably less likely to behave violently than one who identifies strongly with a heroic and attractive James Bond (McCauley, 1998). See Dorr and Kovaric (1980), Tamborini (1991), and Goldstein (1998) for discussions of individual differences in reactions to TV violence.

Arousal. Next, the variable of arousal level of the viewer is important. A person who is already physiologically aroused for whatever reason is more likely to engage in violence after seeing a violent media model than is a nonaroused person (Tannenbaum, 1971, 1980). The arousal may come from the film itself, given that violent films tend to be emotionally arousing and exciting, or it may come from some prior and unrelated source, such as the manipulation in some experiments that makes one group of participants angry before exposing them to a violent media model (e.g., Berkowitz, 1965; Hartmann, 1969; Zillmann, 1978). See Zillmann (1991a) for a review and discussion of the media and arousal issue. This issue of the interaction of arousal and a violent model becomes important when considering sexual violence, which is examined in chapter 10.

Violence: Watching All That Mayhem Really Matters 278

Age and Gender. In terms of age and gender, modeling effects typically increase up to about ages 8 to 12 and slowly decrease thereafter. After this age, children have developed their own viewing set and are better able to separate video experience from reality Although boys consistently both watch more violent TV and are themselves more violent than girls, there is no clear evidence of a stronger modeling effect as such on either boys or girls, at least not before about age 10 (Hearold, 1986).

Context of Violence. One contextual variable that has not yet been studied very extensively is the embedding of graphic violence in a humorous context. For example, violent characters can utter humorous wisecracks, a laugh track can accompany violent scenes, and acts of violence can be presented with humorous consequences. Sometimes technical or editing devices can serve to reinforce violence in such a context. For example, when the hero in the film Natural Born Killers drowned his girlfriend’s father in a fish tank and killed her mother by tying her to a bed, dousing her in gasoline and burning her alive, a background laugh track encouraged us to see this brutality as humorously entertaining, reinforcing positive responses within the viewer. One empirical study of the effects of viewing wisecracking heroes and villains in a violent action film (The Hitman) found a different pattern of results in men and women (King, 2000). The presence of humor in the hero increased distress in women but not in men, while their reactions to

a subsequent nonfiction film showed the reverse effect. There are also a high number of acts of violence on comedy shows, especially in scenes of farce (Potter & Warren, 1998).

This embedding of violence in comedy is a specific instance of the broader situation of violent behavior being reinforced by virtue of its occurrence in a context that is overall very reinforcing. For example, because viewers may choose to identify more with the glamorous opulence of Beverly Hills 90210 than the gritty seediness of NYPD Blue, violence on the former show may have a greater effect, even if the actual violence is less in quantity or graphic specificity. However, this effect may be mitigated by the fact that the more realistic shows may have a greater impact than the less realistic ones due to their greater relationship to the viewers’ own experience, Thus, we see that many factors are at work in determining modeling effects.

We turn now to a final apparent psychological effect of violence: catharsis.

Catharsis

The notion of catharsis extends all the way back to Aristotle’s Poetics, where

he spoke of drama purging the emotions of the audience. In modern times,

279 A Cognitive Psychology of Mass Communication

however, the notion was developed largely in psychoanalytic theory. According to Freud, the id, ego, and superego are locked in battle, with anxiety resulting from id impulses trying to express themselves and, in so doing, coming into conflict with the moralistic superego. The threatening unconscious impulses like sex and aggression are repressed from consciousness but may cause anxiety when they creep back. These repressed impulses and the anxiety that they produce may be dealt with directly by overt sexual or violent behavior or indirectly through some sublimated substitute activity, such as watching others act sexually or violently on television.

The emotional release called catharsis comes from venting the impulse (i.e., expressing it directly or indirectly). This emotional purging has been a notoriously difficult concept to operationally define and test, but it has continued to have a lot of intuitive appeal and anecdotal support (e.g., people report feeling better after watching a scary movie). Catharsis theory does, however, make one very clear prediction about the effect of TV violence on behavior, a prediction that is eminently testable and exactly opposite to the prediction of modeling theory. Whereas modeling predicts an increase in violent behavior after watching media violence, catharsis theory predicts a decrease in such behavior (S. Feshbach, 1955). If the substitute behavior of watching the violence provides the emotional release that would normally result from someone actually acting violently, then violent behavior should decrease after watching media violence. Thus, the two models are clearly and competitively testable. When such tests have been done, modeling theory has been supported (e.g., Siegel, 1956), whereas catharsis theory seldom has. In spite of consistent failures to be supported by scientific evidence over many years (Bushman, Baumeister, & Stack, 1999; Geen & Quanty, 1977; Zillmann, 1978), catharsis continues to occupy a prominent but undeserved place in the conventional wisdom about the effects of media violence.

Later refinements of catharsis theory were proposed (S. Feshbach & Singer, 1971). It may be that the media violence elicits fantasizing by the viewer, and that fantasizing, rather than the media violence per se, is what leads to catharsis. Another version of catharsis theory argues that watching media violence reduces one’s arousal level, and thus one is less prone to violence. There is evidence that a reduction of arousal level is associated with decreased violent behavior. Third, TV violence may elicit an inhibition response, which puts a brake on tendencies toward violent behavior. This is very similar to a sensitization hypothesis. None of these explanations, however, has offered a serious challenge to the overall conclusion that viewing media violence leads to increases in violent behavior.

Violence: Watching All That Mayhem Really Matters 280

WHO WATCHES MEDIA VIOLENCE AND WHY? Another approach to studying media violence has been to examine what

attracts viewers to violence and why some are attracted much more than others (Fenigstein & Heyduk, 1985; Goldstein, 1998; Sparks & Sparks, 2000). What is it that is appealing about violence as entertainment? Some answers that have been suggested are its novelty, sensory delight, and the violation of social norms. It also may have some social utility, for example in allowing the display of mastery of threats or dependence on a loved one (Zillmann & Weaver, 1996).

Social Factors

In an historical sense, very violent films, especially horror films, have generally been the least popular during wartime and have shown very strong popularity during times of overall peace accompanied by high degrees of social unrest.

In terms of gender socialization in adolescents, Zillmann and Weaver (1996) developed and tested a model of the role of social motives in consuming horror films. Specifically, they argued that preadolescent and adolescent boys use horror films to develop mastery over fear and to perfect their displays of fearlessness and protective competence. Girls, on the other hand, use the same films to develop their displays of fearfulness and protective need. While girls actually enjoy the films less than boys do, both find them socially useful as they learn the still very traditional gender roles in dating, whereby the boy is the fearless protector and the girl the dependent and fearful party. The boy’s expression of boredom or amusement in response to graphic violence is thus a statement of his apparent mastery over fear. This mastery then may actually mediate a feeling of pleasure. This pleasure, however, was much greater in the presence of a fearful young female companion than in the presence of a fearless female companion who expressed less dependence. Young men, compared to women, are much more desirous that their date not know how scared they felt while watching a scary movie (Harris, et al., 2000).

Individual Differences

Several researchers have examined the relationship between personality factors and preference for violent media (e.g., Haridakis, 2002). One of the most-studied trait in this regard is empathy. Although empathy is itself a multidimensional construct (M.H.Davis, et al., 1987), it is generally correlated negatively with preference for violent media, especially strongly for the cognitive components like perspective taking and fantasy empathy.

281 A Cognitive Psychology of Mass Communication

Tamborini (1996) has developed a complex cognitive-motivational model of empathy as a predictor of reactions to viewing violence. Empathy may be evoked to different degrees by certain editing techniques and formal features. For example, extended close-ups of faces encourage empathic responses more than rapid-fire wide-angle camera shots, In terms of a path analysis, Raney (2002; see also Raney & Bryant, 2002) argues that the elicitation of empathy may evoke sympathy for the victim of violence, which in turn predicts lesser enjoyment of the violence.

Another personality variable studied in relation to violent media consumption is sensation seeking (Zuckerman, 1994, 1996), the “seeking of varied, novel, complex, and intense sensations and experiences, and the willingness to take physical, social, legal, and financial risks for the sake of such experience” (Zuckerman, 1996, p. 148). Sensation seeking is positively correlated with a preference for viewing violence (Krcmar & Greene, 1999), although the relationship is tempered by the degree of alienation experienced and by the fact that high sensation seekers tend to prefer real-life over vicarious experiences and thus tend to be lighter-than-average TV viewers overall (Slater, 2003).

Another individual difference variable studied is psychoticism. Men (though not women) scoring high in psychoticism found violence more acceptable to solve an interpersonal problem after seeing very violent contemporary action films like Total Recall and Die Hard II, though not after seeing “old-style” violence or horror-film violence (Zillmann & Weaver, 1997). This suggests that personality factors can interact with certain kinds of violence but not others.